Source: ATYUN AI platform
Analysis of startups Descartes Labs today announced that its platform is now out of beta, headquartered in New geospatial Mexico. In the cash-rich company has allowed businesses analysis of satellite images is pulled by NASA and the European Space Agency and the prediction model based on this data, but starting today, weather data it is added, its library, and commercial high-resolution images thanks to the establishment of a new partnership with OneAtlas Airbus project.
As you may recall Descartes laboratory co-founder of Zite Mark Johnson told me that the new team now extract data 100TB regularly every day. Then, the company's customers use these data to predict the growth of crops. Although Descartes laboratory can not disclose most of its customers, but Johnson told me, Cargill and the Los Alamos National Laboratory team is one of its users.
While in theory anyone can access the same data and start analyzing it to thousands of compute nodes and build a model, but as the value of such services is very abstract and all this work for developers and analysts to focus on what they do best thing. .
"If you look at an early beta version of the client system, usually it is a company with some kind of geospatial expertise of the company," Johnson told me. "Typically, they will collect their own data, but the main challenge they face is that their team should spend all the time on scientific research datasets people - most of their time, sometimes more than 80% of the time - they Gathering data, cleaning up the data, the data ready for analysis. Therefore, only a small part of the work time for analysis. "
So far, Descartes laboratory infrastructure, mainly in the Google cloud platform to run, has handled more than 11PB of compressed data. Thanks to cooperation with Airbus, which now also received very high resolution data for its users. For example, some of the resolution Landsat satellite data free of 30 meters per pixel, while Airbus data throughout the world-wide as 1.5 meters per pixel, in 2,600 cities in 50 cm per pixel. Adding to this global NOAA's weather data, it is easy to imagine what a developer can build a model based on all the information.
Johnson told me that many users will bring their own data services to build better models.
Although Descartes laboratory focuses on the early developers, but it is worth noting that the team now has to build a viewer that allows any user (to pay for services) using the base map and add more information at the top.
Johnson told me that the team plans to add more data sets over time, although the focus of the service will always remain on the spatial data.
This switched ATYUN artificial intelligence media platforms, the original link: Descartes laboratory launched its Geospatial Analysis Platform
more recommendations
Cybercriminals machine learning
Drop by gradient reverse engineering, acquisition of various constituents of food
Technology makes life better use of artificial intelligence to predict urban flooding
AI need big data, and large data needs to AI
